How fast before step loss

[creep_pea]

Hi

Just wondering how fast my boss can go around corners without loosing steps.

Is there a recommended limit? as my machine looses steps if run flat out on small rads, or am I best just finding a safe limit with trial and error?

Cheers

Chris

[machintek]

This is a compound problem. First of all if this is a BOSS 5, it step no finer than .001 inches. This will not make for a smooth circle and at the same time may be part of the mispositioning issue. It may not think in small enough steps. A BOSS 6 steps in .0005 inches. A bit better.
My BOSS 9 had servo motors so it could create a lot smoother curves and it had a G code that would speed up the feed rate around an outside radius and slow it down on a inside radius. I do not believe this G code existed prior to the BOSS 8.

Then there is the strength of the motors and the mass of the table and saddle whose direction you are trying to change. This is basic physics. Can the motor and drive accomplish this in the time and space you want it to?

What is your definition of "flat out"? Rapid? Or a feedrate?

George

[creep_pea]

My machine is a BOSS 6.1 so it steps in .0005 inches and the smoothness of the rads is not a problem when running at a slow enough feedrate.
I'm making a prototype job out of wood at the moment so I was running the feedrate upto it maximum of 1300mm/min (51ipm) but the step loss was visable (as much as a couple of millimeters) so I'm turning the feedrate down to around 30% to go around the corners at the moment.
I guess there is no definate answer to how fast it will go without step loss and I'll have to find a reasonable feedrate that works.

Thanks

Chris

[Hood]

Chris, get that thing retro fitted with modern steppers, you wont regret it, 3 or 4M/min should be no problem
Hood

[creep_pea]

Hood

If it was making me a bit of money I might consider it, not making a penny at the moment.
Are you still using stepper motors on yours then, though you be running servo's with encoders.

Chris

[Hood]

Chris
Still not got the BOSS finished yet, I could have it completed in less than a day but I am waiting until I get into the new workshop before I do that. The reason is that I have the BOSS stored in another building at the harbour and to move it into my present workshop would be a lot of hassle just for a month or so (hopefully, been saying that for two or three months already) The builders have now finished laying the cobbles outside the new workshop so I can get access but in a week or so the council are away to dig up right at the entrance gate to fit a new drain so I wouldn't be able to get in very easily, so I will hang off a while yet.
I have steppers on my manual mill and they have been excellent so far, have read all the pros and cons for steppers v servos and in the end decided that the price difference wasnt worth it for the increased speed that servos would give me (dont think I really need that speed). Also steppers are only £270 for the three, servos would be that (and probably more ) each.
Hood

[jdelaney44]

Well, now that I have a working machine I can comment from a more educated position. If there is no way to stop the controller from running when the servos fault, there is almost no use to servos. I have servos using Gecko G320s now.

Yeah, the Geckos fault out, but then they wait a few seconds and retry. Meanwhile back at the PC Mach3 is merilly sending out steps like nothing ever happenned 'cause there's no feedback loop. This makes stepping away from the machine for 5 seconds a little dicey.

So in the event of a tool crash or lost steps cause you tried a 3 axis rapid with out enough juice to back it up you just lost your datum and you are now milling a door stop. An ugly one at that I suspect.

So servos with no feedback to the control software are no better than steppers. Pick your poison I s'pose.

Now then, I think the GREX is going to help solve this and there is another JT Cullins card out there that at least stops Mach3.

I know this doesn't address a BOSS controller, but I thought I'd add my nickel to the stepper / servo fray.

With EMC as the controller and a Pico systems card, I understand you can get real servo performance that adapts feed rate when steps start going away. Could be wrong on that.

I would also hazard a guess that this is really a bigger issue than people are making it out to be.

I also think there is a way to hijack the test point voltage change on the Geckos and use a key board emulator to ramp down Mach3's feed rate....Hmmm. That's a whole 'nother Oprah.

-jd

[jdelaney44]

Just checked and the Gecko GRex set up should be coming soon. As well as the Mach4 support for it. That should solve the feedback issues I think.

-jd

[jdelaney44]

Correction. The G320 retries after 3 seconds if you have the err terminal jumpered to the encoder supply terminal. This is how I am set. This could probably be used to trip and e-stop in Mach3 through the sound logic board. I'm going to look into this for a mean time fix.

Sorry.......